Effect of nitridation on the aqueous dissolution of
Na2O-K2O-CaO-P2O5 metaphosphate glasses
Quentin Riguidel, Francisco Muñoz*
Instituto de Cerámica y Vidrio (CSIC), Kelsen 5, 28049 Madrid (Spain)
Abstract
The use of oxynitride glasses is presented as an alternative for the preparation of
bioresorbable phosphate glasses with a controlled dissolution rate. This work describes
the design of oxynitride phosphate glasses within the systems of composition (50x)Na2O.xCaO.50P2O5 and (25-(x/2))Na2O.(25-(x/2))K2O.xCaO.50P2O5 (x=5,10,15,20
mol %) throughout the processing parameters of the ammonolysis reaction and the glass
composition. Mixed-alkali sodium-potassium phosphate glasses with low CaO contents
are those presenting the most adequate characteristics for their nitridation. The
dissolution rate has been determined by immersion of glass samples in water, at
constant temperature of 37ºC, and it has been discussed as a function of both modifiers
composition and nitrogen content incorporated in the glasses through ammonolysis. All
oxynitride glass compositions dissolve congruently and their dissolution rate decreases
by more than three orders of magnitude for the highest nitrogen contents. However, it
has been demonstrated that nitrogen contents as low as 2-3 wt. % (i.e. 0.2 N/P ratio) are
sufficient to decrease the dissolution rate by one order of magnitude with respect to the
pure oxide glasses. Novel oxynitride phosphate glasses with a controlled and congruent
dissolution are proposed for future applications in biodegradable composite materials,
tissue engineering or host matrices for the controlled release of drugs.
Keywords: Phosphate Glasses; Bioresorbable Glasses; Oxynitride Glasses; Controlled
1
release
(*) Corresponding author:
Dr. Francisco Muñoz
Tel : +34917355840 (ext. 1225)
Fax : +34917355843
e-mail : [email protected]
2
1. Introduction
Phosphate based glasses are a special kind of materials whose composition and
properties can be tailored within a very wide range. Generally, phosphate glasses posses
low glass transition temperature, high coefficient of thermal expansion and high
transmission to the ultraviolet wavelengths. Furthermore, their low melting temperature
and special chemistry contribute to produce glasses which can host relatively high
amounts of transition metal ions as well as rare earth elements. To date, many potential
applications have been proposed for phosphate glasses. Their special thermal behaviour
makes them suitable for low temperature sealing materials and numerous studies have
been published on it [1-4]. Nd-doped phosphate glasses have received special attention
for their application as solid-state laser host matrices with exceptional performance [5].
Feasibility for the application of phosphate glasses and protonic conductors [6], nuclear
waste host matrices when doped with iron or lead oxides [7] as well as semiconducting
glasses [8] have also attracted much attention. More recently, phosphate glasses have
been studied for their interest within the Biomaterials field. Many authors have studied
the dissolution properties of phosphate glasses and their potential use as bioresorbable
materials and composites [9-12]. In particular, Knowles remarked the “ability to
dissolve completely in aqueous media” of phosphate glasses, as a remarkable feature
that can be exploited from the biomedical point of view [13]. He also reminds that the
adequate modification of their composition can give rise to bioglasses with sufficient
chemical durability and perfect biocompatibility with the ions found in the body. Thus,
phosphate glasses may offer several interesting applications [14], such as glass-polymer
composites for tissue engineering [15], antimicrobial agents when the glasses are doped
with functional elements like Ga [16], or in the form of glass fibres [17,18].
However, the most important factor limiting the practical application of phosphate
3
glasses is their extremely low chemical durability. A demonstrated way to radically
improve the chemical durability of phosphate glasses is based on the partial substitution
of nitrogen for oxygen, firstly developed by Marchand in the early 80’s [19].
The structure of phosphate glasses is built up of PO4 tetrahedra, which are named as Qn
groups, with n the number of bridging oxygen atoms, according to Lippmaa et al. [20].
Modifier ions are bonded to non-bridging oxygens and, depending on their
concentration, phosphate glasses are classified in ultraphosphate for O/P < 3;
metaphosphates for O/P = 3; and polyphosphate glasses for O/P > 3. In oxynitride
glasses, nitrogen substitutes both bridging and non-bridging oxygens and it can appear
in the form of dicoordinated, Nd (-N=), or tricoordinated, Nt (-N<) species, where the
Nd/Nt ratio also depends on the initial composition of the glasses [21]. For a
metaphosphate glass with composition MPO3 (M=modifier), corresponding oxynitride
glasses can be formulated as MPO3-3x/2Nx, due to the fact that 2 N3- substitute for 3 O2-,
and nitrogen content has been usually expressed as the N/P ratio. For all nitrogen
contents there is a simultaneous increase in the amount of both Nt and Nd, however, the
proportion of Nt is slightly higher than that of Nd except for the highest N/P ratios, as
seen by XPS [21]. Simultaneously, new PO3N and PO2N2 tetrahedra are formed from
PO4 ones [21], which can comprise both Nt and Nd nitrogen species in the same
tetrahedra. The increased bonding density and higher covalent character of the P-N
bonds against P-O ones is responsible for the drastic increase in chemical durability of
the glasses. The dissolution mechanism in phosphate glasses takes place throughout the
hydration and release of phosphate chains. However, it has been proposed that the
dissolution in oxynitride glasses must proceed through the hydrolysis of P-N bonds, due
to the much lower hydration tendency of the oxynitride tetrahedra, which constitutes a
big hindrance for their dissolution [22]. In metaphosphate glasses, all phosphate chains
4
turn to be highly interconnected through new P-N bonds, thus dissolution of the
polymeric chains is prevented.
Most of the works on oxynitride phosphate glasses have been focused on the kinetics of
nitridation [23], the structure of the oxynitride glass network [24], and the influence of
nitrogen on the main properties of the glasses [25]. The introduction of nitrogen might
extend the applicability of phosphate glasses and some other potential uses of
oxynitrides have been proposed, like their use as all-solid-state electrolytes for lithium
rechargeable batteries due to the increased electrical conductivity of the oxynitride glass
[26].
The low chemical durability of bioresorbable phosphate glasses can also be improved
by addition of intermediate elements, e.g. Al, Fe, Ti, though these elements can
significantly alter other properties of the glasses, such as their thermal behaviour, thus
being not always the best solution. In this work, an alternative for the preparation of
bioresorbable phosphate glasses with a controlled dissolution rate is present on the basis
of the nitrogen for oxygen substitution in metaphosphate glasses with composition (50x)Na2O.xCaO.50P2O5 and (25-x/2)Na2O.(25-x/2)K2O.xCaO.50P2O5. The synthesis of
the oxynitride glasses is described throughout the processing parameters involved
during the ammonolysis reaction as well as the influence of composition on the
suitability of the glasses for nitridation. The dissolution rate has been determined by
immersion in water at constant temperature of 37ºC and it has been discussed as a
function of both modifiers composition and nitrogen content incorporated in the glasses
after ammonolysis.
2. Materials and methods
Metaphosphate base glasses with compositions (50-x)Na2O.xCaO.50P2O5 and (25-
5
(x/2))Na2O.(25-(x/2))K2O.xCaO.50P2O5 (5<x<20 mol %) have been obtained from
mixtures of reagent grade NaH2PO4.H2O (Merck, 98%), CaCO3 (Panreac, 99%), K2CO3
(Panreac, 98%) and NH4H2PO4 (Aldrich, 97%). Batches for 100 g glass were calcined at
400ºC and melted in air at 800ºC during 2 h using porcelain crucibles, then poured onto
a brass plate. The glasses were annealed above their glass transition temperature and
their amorphous nature confirmed by X-ray diffraction.
Oxynitride phosphate glasses were obtained through ammonolysis of the base glasses
inside an Al2O3 gas-tight tube furnace at 700ºC and 750°C, during times from 1 to 8 h.
A rectangular graphite plate with up to 8 cavities (=2.5 cm and 0.3 cm depth) for
containing the base glass samples (2-3 g) was used. The furnace is heated up to the
treatment temperature at a constant heating rate of 10 K min-1 under N2 flow. Then, the
nitrogen flow is switched to NH3 (<400 ppm H2O) during the ammonolysis treatment
time, and switched back to nitrogen while leaving the furnace to cool down to room
temperature. Nitrided glasses presented a light greyish colour due to graphite surface
contamination in the base of the glass in contact with the graphite plate, which can be
totally removed after polishing.
Glass transition temperature (Tg) was determined by means of Differential Thermal
Analysis (DTA) in a SEIKO 6300 analyser, using platinum crucibles and a constant
heating rate of 10ºC.min-1, within the temperature range from 25 to 800ºC under air
flow.
Nitrogen contents in oxynitride glasses were determined using the inert gas fusion
method in a differential N2/O2 LECO TC-436 analyser. Nitrogen content is obtained as
wt. % in the glass, but it will be expressed as the N/P ratio according to the glass
formulations Na2(0.5-y)CayPO3-3x/2Nx and Na2(0.25-y)K2(0.25-y)CayPO3-3x/2Nx (y=0.05, 0.1,
0.15, 0.2) for both systems of composition. At least three analyses of N2 were done in
6
the same conditions for each sample, after which n-1 (N/P ratio) = 0.02.
X-Ray Diffraction (XRD) analysis of the base glasses was carried out with a Bruker D-8
diffractometer using monochromatic Cu Kα radiation (1.5418 Å) between 10º and 70º in
2.
The density of the base glasses was determined by helium pycnometry in a
Quantachrome Corp. multipycnometer by using bulk glass samples.
Prismatic samples, with dimensions approximately 1x0.5x0.5 (in cm) and surface area
of between 2 and 3 cm2, were cut and polished (with 2500-grit SiC paper) from
annealed base glasses as well as from oxynitride ones. Sample mass was about 0.5 g.
The dissolution rate tests were performed by measuring the weight loss in distilled
water with pH = 6  0.2, after 24 h of immersion. The samples were immersed into
plastic bottles and placed on top of a chair-like platinum wire. A constant surface area to
water volume of 0.04 was used for all tests. Prior to dissolution experiments glass
samples were cleaned with acetone in an ultrasonic bath and dried before weighing.
After dissolution experiments the tested samples were dried until constant weight (
1.10-4 g) and the dissolution rate (Dr) was calculated from the weight loss (m, in g)
normalized to the initial glass surface area (S, in cm2) and dissolution time (t, in h), for a
constant immersion period of 24 h, according to the equation:
Dr = m/(S.t)
Chemical
analyses
of
the
dissolution
(1)
tests
in
glass
with
composition
35Na2O.15CaO.50P2O5 and oxynitride glasses for different nitrogen contents, were
performed through Inductively Coupled Plasma-Emission Spectrometry (for Ca and P),
in a Thermo Jarrel Ash IRIS Advantage equipment, and Flame Photometry (for Na) in a
Perkin Elmer 2100 instrument. Total error of analysis is estimated to be within  1%.
7
3. Results and discussion
Table I gathers the composition, glass transition (Tg), melting temperature (Tm), density
and molar volume of phosphate glasses of the series (50-x)Na2O.xCaO.50P2O5 and (25x/2)Na2O.(25-x/2)K2O.xCaO.50P2O5 (x=5,10,15,20). Molar volume of the glasses (Vm)
has been calculated from density measurements by using the equation:
Vm (in cm-3.mol-1) = M/d
(2)
being M the molecular mass and d the density of the glasses. For the two glass systems,
it can be seen that Tg increases with increasing CaO content, which is due to the higher
Ionic Field Strength (IFS) of Ca2+ ions compared to those of Na+ and K+ ones, being IFS
equal to 0.33, 0.19 and 0.13 for Ca2+, Na+ and K+, respectively, according to Dietzel
[27]. Furthermore, Tg of the mixed-alkali system, Na2O-K2O-CaO-P2O5 glasses, is
about 40 to 60ºC lower than those of the Na2O-CaO-P2O5 system. It is also observed
that melting temperature increase with CaO content being also smaller for the K2Ocontaining glasses (Fig. 2). Molar volume of Na2O-K2O-CaO-P2O5 glasses shows a
slight decrease with CaO content, but no change in the Na2O-CaO-P2O5 system. Again,
the higher IFS of Ca2+ ions, compared to Na+ and K+, produces a contraction effect of
the glass network resulting in slightly lower molar volume, which is a little more
pronounced in Na2O-K2O-CaO-P2O5 glasses due to the even lower IFS of K+.
While the glass transition does not have a direct influence for the nitridation of
phosphate glasses, the melting temperature of the crystalline phases formed by
devitrification might play an important role. Nitridation has to be carried out in the melt,
i.e. above Tm. If the reaction is performed either below or very close to the melting
temperature, the crystallisation of the glass during heating, preceding the reaction with
ammonia, or during cooling, once the reaction has finished, might lead to a ceramised
material preventing any proper nitridation. On the other hand, maximum reaction
8
temperature is limited to 800ºC due to the possibility of phosphorous reduction which
would alter the composition and homogeneity of the melt. Then, nitridation
temperatures of 700 and 750ºC have been chosen for this study, and short times up to 8
h, thus being not very severe conditions. As it will be discussed below, nitrogen contents
of about 2 to 3 wt. % in the glasses are enough to greatly improve the chemical
resistance to hydrolysis, therefore temperatures as low as 700ºC during a few hours of
treatment are suitable to get such a nitrogen contents. Melting temperatures of Na2OCaO-P2O5 glasses are just below 700ºC for 5 mol % CaO composition and between 700
and 750ºC for  10 mol % CaO. If nitridation were performed at 700ºC, deterioration of
the glass could takes place during either the reaction with ammonia or cooling.
However, it has been observed that increasing CaO helps to carry out the reaction
without devitrification phenomena. Calcium is known to be a stabiliser of the glass
network due to its higher IFS, and Ca-containing glasses will also have a higher
viscosity than pure alkali glasses, which is reflected in the lowest tendency for
crystallisation of the calcium phosphate glasses for increasing mol % CaO. Glasses for
5 and 10 mol % CaO present clear devitrification after nitridation treatment, while 15
and 20 mol % CaO developed very light or no crystallisation, giving rise to an
homogeneous glass. Thus, under these conditions, the higher the CaO content the lowest
the tendency for devitrification. On the other hand, oxynitride glasses of the Na2O-K2OCaO-P2O5 mixed-alkali system did not present any evidence of crystallisation after
treatments. Melting temperature of 5 and 10 mol % CaO K2O-containing glasses are
lower than 700ºC, thus devitrification is not expected to influence during or after
nitridation. Furthermore, the higher the CaO content the higher the viscosity of the melt
at the treatment temperature, which can help to prevent devitrification phenomena
independently of the nitrogen content achieved.
9
3.1. Nitrogen incorporation
The nitrogen for oxygen substitution has been studied at 700ºC and 750ºC, for reaction
times between 1 and 8 h, in both Na2O-CaO-P2O5 and Na2O-K2O-CaO-P2O5 glasses. As
it has been shown before, Na2O-CaO-P2O5 glasses presented clear devitrification for
most compositions with the lowest CaO content and, especially, low nitridation times.
On the other hand, the increase in CaO content should give rise to a melt with higher
viscosity. Therefore, nitridation of a higher viscosity melt resulted sometimes in glasses
with presence of bubbles. Particularly, Na2O-CaO-P2O5 glasses were those with more
bubbling problems at high CaO concentration.
Figures 1 and 2 depict the nitrogen content as a function of the reaction time for
increasing CaO contents
in
(25-x)Na2O.xCaO.50P2O5
and (25-x/2)Na2O.(25-
x/2)K2O.xCaO.50P2O5 glass systems, respectively. As pointed out above, nitrogen is
expressed as the N/P ratio following the formulation Na2(0.5-y)CayPO3-3x/2Nx and Na2(0.25y)K2(0.25-y)CayPO3-3x/2Nx
(y=0.05, 0.1, 0.15, 0.2). The nitrogen content increases with the
reaction time in a similar way to that seen in other phosphate glass systems. First,
nitrogen increases following a linear function of the square root of time [23] up to a
maximum which depends on the processing conditions as well as on temperature.
Meanwhile Na2O-CaO-P2O5 glasses show approximately the same values of N/P ratio
for all CaO formulations, K2O-containing glasses clearly present smaller nitrogen
contents for 15 and 20 mol % CaO-containing glasses. It is also important to note that in
both systems, glasses with the highest CaO content do not allow the preparation of
bubble-free homogeneous glasses and nitrogen content can vary in a big extent, thus no
suitable data have been obtained to be used for dissolution tests. Slightly higher
nitrogen contents are also found in the Na2O-CaO-P2O5 glasses.
10
Figure 3 shows the nitrogen content as a function of the reaction time at 700ºC and
750ºC in 17.5Na2O.17.5K2O.15CaO.50P2O5 glass, showing the kinetics of nitrogen
incorporation for times up to 6 h treatment. Nitrogen content is up to  0.2 units N/P (2
wt.%) higher when nitridation is carried out at 750ºC than at 700ºC. This behaviour can
be explained through the different viscosity of the melt at the reaction temperature. It
was reported by Muñoz et al. that nitrogen content decreases linearly with the increase
in viscosity of the melt at a constant reaction temperature [23], thus the higher the
viscosity the lower the nitrogen incorporation. This is reflected in the lower amount of
nitrogen found in glasses nitrided at 700ºC with those at 750ºC. As seen in Fig. 2,
nitrogen decreases for higher CaO contents in nitrided Na2O-K2O-CaO-P2O5 glasses,
which can also be due to the increased viscosity of the CaO-containing glasses. The
higher the amount of CaO the higher the viscosity of the melt, i.e. less nitrogen can be
incorporated.
3.2. Hydrolytic resistance of oxynitride glasses
Figures 4 and 5 present the Log of the dissolution rate (Dr), normalised to the sample
surface area and time of immersion in water at 37ºC after a 24 h test, as a function of the
N/P ratio in oxynitride glasses of the systems of composition Na2O-CaO-P2O5 and
Na2O-K2O-CaO-P2O5, respectively. The choice for representing dissolution rate vs
nitrogen in logarithmic scale has only been made to display the wide range of data
points. Several common features can be observed in both Figures. First and most
important one, the higher the nitrogen content the lower the dissolution rate. It is worth
pointing out that Log Dr can be up to 3 orders of magnitude lower for the highest
nitrogen contents in most glass compositions. Oxynitride glasses from base composition
22.5Na2O.22.5K2O.5CaO.50P2O5 are those presenting the smallest decrease of
11
dissolution rate for all nitrogen contents. Furthermore, for the 24 h dissolution test in
water, the dissolution rate can only be determined for 15 and 20 mol % CaO nonnitrided phosphate glasses. Low amount of CaO in phosphate glasses results in their
dissolution much before 24 h after immersion in water due to their extremely high
dissolution rate. Then, the higher the CaO content the lower the dissolution rate of the
phosphate glass. Thus, it is clearly seen the effect of both calcium and nitrogen on the
dissolution of the phosphate and oxynitride glasses. It was also observed that higher
nitrogen contents are obtained for the lower CaO contents. However, similar dissolution
rates can be attained for low [CaO] and high [N2] and vice versa, especially for the
glasses of the system Na2O-CaO-P2O5. This means that nitrogen content can be varied
according to the original glass composition to find the adequate dissolution behaviour.
The experimental data points have been fitted to a linear equation of Log Dr against the
N/P ratio, and an approximate linear decrease of the dissolution rate with nitrogen is
observed in all cases. However, it must be taken into account that the error in the
determination of the dissolution rate through the weight difference of the sample before
and after dissolution can be very large, hence the high dispersion of the points around
the linear fits. Some of the experimental sources for the dispersion might come from the
fact that the possible saturation of dissolution products near the sample surface as well
as the variation of the sample surface area with dissolution time could not been
considered.
Figure 6 depicts a comparison of the dissolution at 37ºC in the oxynitride glasses
belonging to the 10 mol % CaO Na2O-CaO-P2O5 and Na2O-K2O-CaO-P2O5 systems. It
is clearly seen that, independently of the glass system, the oxynitride glasses can subject
of a very fast decrease in their dissolution rate for only N/P=0.2, i.e. up to 10 times
lower than the dissolution rate in the oxide glasses. The dissolution rate continues
12
decreasing for all nitrogen contents. Low CaO contents allow introduction of a higher
amount of nitrogen and better glass-forming ability of the corresponding oxynitride
glasses. However, independently of the glass composition, it is remarkable that only
moderate nitrogen contents are sufficient to ensure a relatively low dissolution rate.
Table II gathers the concentration of oxides (in mol %) and the P2O5/(Na2O+CaO) ratio,
analysed in the dissolutions of the 35Na2O.15CaO.50P2O5 base glass and oxynitride
glasses for different amounts of nitrogen after the dissolution tests. The comparison of
the concentration of all oxide components in the resulting dissolutions with the
concentration in the nominal composition of the same base glass are within  1 % for all
nitrogen contents. This is representative of a congruent dissolution of all components of
the glasses with no influence of the nitrogen content. Then, the oxynitride glasses
dissolved congruently and without formation of any precipitated layer by accumulation
of dissolution products onto the sample surface.
For the two glass systems under study, it has been seen that an adequate combination of
both glass composition and ammonolysis processing parameters can give rise to low or
moderate nitrogen containing oxynitride phosphate glasses with controlled and
congruent dissolution. Numerous compositions, especially alkali and alkaline-earth
phosphate glasses, could be nitrided with the purpose of being used as biocompatible
materials for tissue engineering, in the form of either bulk or fibre glasses, or host
matrices for the controlled release of drugs and antibacterial elements, for instance. In
order to ensure a future potential application as biomaterials, further studies should be
made of the dissolution rate of oxynitride phosphate glasses within a simulated body
fluid as well as on the effect of nitrogen with the biological tissues.
13
Conclusions
In the present work, the influence of composition on the nitridation and dissolution rate
of Na2O-CaO-P2O5 and Na2O-K2O-CaO-P2O5 glasses has been studied. Two major
effects are worth mentioning: first, mixed-alkali sodium/potassium phosphate glasses
allow ammonolysis at lower temperatures, likely due to the lower viscosity of their
melts; and second, increasing CaO content diminishes the tendency for devitrification of
the glasses after the ammonolysis reaction. However, very high amount of CaO gives
rise to glasses which present fining problems. Thus, the mixed alkali system together
with relatively low amounts of CaO constitutes the best compositional choice for
nitridation.
It has been observed that Na and Na/K CaO-containing oxynitride phosphate glasses
dissolve congruently. Their dissolution rate decreases by one order of magnitude for
nitrogen contents as low as N/P=0.2 (3 wt.%), and it continues decreasing with nitrogen
content incorporated in the glasses. It is then suggested that novel oxynitride phosphate
glasses with a controlled dissolution rate, and no necessity of transition metal ions,
could be envisaged for future applications in biodegradable materials, tissue engineering
or host matrices for the controlled release of drugs.
Acknowledgments
Q.R. thanks support from IUT-Chimie de Rennes for his research training stage at the
ICV-CSIC. Discussions from Prof. A. Durán are greatly acknowledged.
14
References
[1] Frieser RG. A review of solder glasses. Electrocomponent Sci. Technol. 1975;2:163199.
[2] Chambers RS, Gerstle Jr. FP, Monroe SL. Viscoelastic effects in a phosphate glassmetal seal. J. Am. Ceram. Soc. 1989;72:929-932.
[3] He Y, Day DE. Development of a low temperature phosphate sealing glass. Glass
Technol. 1992;33:214-219.
[4] Ray NH, Plaisted RJ, Robinson WD. Oxide glasses of very low softening point. Part
4. Preparation and properties of ultraphosphate glasses containing boric oxide. Glass
Technol. 1976;17:66-71.
[5] Campbell JH, Suratwala TI. Nd-doped phosphate glasses for high-energy/high-peakpower lasers. J. Non-Cryst. Solids 2000;263&264:318-341.
[6] Abe Y, Hayashi M, Iwamoto T, Sumi H, Hench LL. Superprotonic conducting
glasses containing water. J. Non-Cryst. Solids 2005;351:2141-2144.
[7] Reis ST, Karabulut M, Day DE. Structural features and properties of lead-ironphosphate nuclear wasteforms. J. Nuclear Mat. 2002;304:87-95.
[8] Murawski L, Chung CH, Mackenzie JD. Electrical properties of semiconducting
oxide glasses. J. Non-Cryst. Solids 1979;32:91-104.
[9] Bunker BC, Arnold GW, Wilder JA. Phosphate-Glass Dissolution in AqueousSolutions. Journal of Non-Crystalline Solids 1984;64:291-316.
[10] Navarro M, Ginebra MP, Clément J, Martínez S, Avila G, Planell JA. Physicochemical degradation of resorbable phosphate glasses stabilized with TiO2. J. Amer.
Ceram. Soc. 2003;86[8]:1345-52.
[11] Vogel P, Wange P, Hartmann P. Phosphate glasses and glass-ceramics for medical
applications. Glasstech. Ber. Glass. Sci. Tech. 1997;70:220
15
[12] Navarro M, Amaral I, Engel E, Barbosa M, Ginebra MP, Planell JA. Fully
biodegradable PLA/CaP glass composite material for bone regeneration: surface
properties and in vitro cell response. Acta Biomaterialia 2005;1:411-419.
[13] Knowles JC. Phosphate based glasses for biomedical applications. J. Mat. Chem.
2003;13:2395-2401.
[14] Abou-Neel EA, Pickup DM, Valappil SP, Newport RJ, Knowles JC. Bioactive
functional materials: a perspective on phosphate-based glasses. J. Mat. Chem. 2009;
19:690-701.
[15] Prabhakar RL, Brocchini S, Knowles JC. Effect of glass composition on the
degradation
properties
and
ion
release
characteristics
of
phosphate
glass-
polycaprolactone composites. Biomaterials 2005;26:2209-2218.
[16] Valappil SP, Ready D, Abou-Neel EA, Pickup DM, Chrzanowski W, O’Dell LA,
Newport RJ, Smith ME, Wilson M, Knowles JC. Antimicrobial Gallium-Doped
Phosphate-Based Glasses. Ad. Func. Mat. 2008;18:732-741.
[17] Lin ST, Krebs SL, Kadiyala S, Leong KW, LaCourse WC, Kumar B. Development
of bioabsorbable glass fibres. Biomaterials 1994;15:1057-1061.
[18] Abou-Neel EA, Young AM, Nazhat SN, Nowles JC. A facile synthesis route to
prepare microtubes from phosphate glass fibres. Ad. Mat. 2007;19:2856-2862.
[19] Marchand R, Nitrogen-containing phosphate glasses, J. Non-Cryst. Solids
1983;56:173-178.
[20] Lippmaa E, Magi M, Samoson A, Engelhardt G. Structural studies of silicates by
solid-state high-resolution silicon-29 nmr. J. Am. Chem. Soc. 1980;102:4889-4893.
[21] Muñoz F, Pascual L, Durán A, Montagne L, Berjoan R, Palavit G, Marchand R.
Structural study of phosphorus oxynitride glasses LiNaPbPON by nuclear magnetic
resonance and X-ray photoelectron spectroscopy. J. Non-Cryst. Solids 2003;324:142-
16
149.
[22] Muñoz F, Durán A, Marchand R, Pascual L. Corrosion behaviour of alkali-lead
oxynitride metaphosphate glasses in acidic medium. Phys. Chem. Glasses 2003;44:416421.
[23] Muñoz F, Pascual L, Durán A, Marchand R. Compositional and viscosity influence
on the nitrogen/oxygen substitution reactions in phosphate melts. Phys. Chem. Glasses
2005;46:39-45.
[24] Marchand R, Agliz D, Boukbir L, Quemerais A. Characterization of nitrogen
containing phosphate glasses by x-ray photoelectron spectroscopy. J. Non-Cryst. Solids
1988;103:35-44.
[25] Reidmeyer MR, Day DE, Phosphorus oxynitride glasses. J. Non-Cryst. Solids
1995;181:201-214.
[26] Yu X, Bates JB, Jellison Jr. GE, Hart FX. A stable thin-film lithium electrolyte:
lithium phosphorus oxynitride. J. Electrochem. Soc. 1997;144: 524-532.
[27]
Dietzel
AZ,
Die
Kationenfeldstärken
und
ihre
Beziehungen
zu
Entglasungsvorgängen, zur Verbindungsbildung und zu den Schmelzpunkten von
Silicaten. Z. Elektrochem. 1942;48:9-23.
17
Figure captions
Figure 1. Nitrogen content (in N/P ratio) as a function of ammonolysis time at 700ºC in
glasses (50-x)Na2O.xCaO.50P2O5 (x=5,10,15,20 mol %). Lines have been drawn as a
guide for the eyes.
Figure 2. Nitrogen content (in N/P ratio) as a function of ammonolysis time at 700ºC in
glasses (25-x/2)Na2O.(25-x/2)K2O.xCaO.50P2O5 (x=5,10,15,20 mol %). Lines have
been drawn as a guide for the eyes.
Figure 3. Nitrogen content (in N/P ratio) as a function of ammonolysis time at 700ºC
and 750ºC in glass with composition 17.5Na2O.17.5K2O.15CaO.50P2O5. Lines have
been drawn as a guide for the eyes.
Figure 4. Log of dissolution rate (Dr), in g.cm-2.h-1, as a function of the nitrogen content
in glasses of the system (50-x)Na2O.xCaO.50P2O5 (x=5,10,15,20 mol %) The
experiments have been performed at 37ºC during a period of 24 h. Lines are least
squares fits of Log Dr vs nitrogen content.
Figure 5. Log of dissolution rate (Dr), in g.cm-2.h-1, as a function of the nitrogen content
in glasses of the system (25-x/2)Na2O.(25-x/2)K2O.xCaO.50P2O5 (x=5,10,15,20 mol
%). The experiments have been performed at 37ºC during a period of 24 h. Lines are
least squares fits of Log Dr vs nitrogen content.
Figure 6. Comparison of the dissolution rate, in g.cm-2.h-1, against the nitrogen content
for
the
10
mol
%
CaO
glasses,
40Na2O.10CaO.50P2O5
and
20Na2O.20K2O.10CaO.50P2O5. The experiments have been performed at 37ºC during a
period of 24 h. Lines have been drawn as a guide for the eyes.
18
Table captions
Table I. Glass transition temperature (Tg), maximum melting temperature (Tm), density
and molar volume (Vm) of the investigated base glasses, (50-x)Na2O.xCaO.50P2O5 and
(25-x/2)Na2O.(25-x/2)K2O.xCaO.50P2O5 (x=5,10,15,20 mol %).
Table II. Analysed values of oxides in dissolution (in mol %) and P2O5/(Na2O+CaO)
ratio of glass 35Na2O.15CaO.50P2O5 and corresponding oxynitride glasses for different
nitrogen contents (in N/P ratio).
19
Table I
Glass
Tg ( 3°C)
Tm (ºC)
Density Molar Volume
(g.cm-3) (cm3.mol-1 )
45Na2O.5CaO.50P2O5
296
695
2.51
40.53
40Na2O.10CaO.50P2O5
305
726
2.59
39.17
35Na2O.15CaO.50P2O5
318
739
2.52
40.15
30Na2O.20CaO.50P2O5
338
738
2.52
40.14
22.5Na2O.22.5K2O.5CaO.50P2O5
239
667
2.47
44.04
20Na2O.20K2O.10CaO.50P2O5
263
693
2.50
43.19
17.5Na2O.17.5K2O.15CaO.50P2O5
278
703
2.49
42.88
15Na2O.15K2O.20CaO.50P2O5
300
703
2.48
42.55
20
Table II
Oxide in glass
N/P = 0
35Na2O.15CaO.50P2O5 (base glass)
N/P = 0.16
N/P = 0.25 N/P = 0.35
Na2O
32
36
34
33
CaO
17
14
16
17
P2O5
51
50
50
50
P2O5/(Na2O+CaO)
1.04
1
10
1
21